Introduction

Achieving glycemic control is the main objective in the treatment of patients with type 2 diabetes (T2D), with a glycated hemoglobin (HbA1c) target of < 7% recommended for most adults [1, 2]. If HbA1c targets are not reached after initiating basal insulin therapy in patients with T2D, the American Diabetes Association (ADA) guidelines suggest considering a combination injectable therapy, such as rapid-acting insulin prior to the largest meal, glucagon-like peptide-1 receptor agonists (GLP-1 RAs), or a switch to premixed insulin twice daily [2].

iGlarLixi is a once-daily, titratable, fixed-ratio combination of insulin glargine U100 (iGlar) and the GLP-1 RA lixisenatide (Lixi). The complementary actions of iGlar, which predominantly targets fasting plasma glucose (FPG), and Lixi, which predominantly targets postprandial plasma glucose levels, may benefit patients with T2D who are unable to achieve their glycemic targets [3,4,5]. iGlarLixi was approved in the USA in 2016 for the treatment of adults with T2D inadequately controlled on basal insulin (< 60 U/day) or Lixi, and in Europe in 2017, in combination with metformin, for adults with T2D inadequately controlled with metformin alone or metformin combined with another oral anti-diabetes drug (OAD) or with basal insulin [6, 7]. iGlarLixi has demonstrated superior reduction in HbA1c compared with its individual components of iGlar and Lixi, in the LixiLan-O, LixiLan-L, and LixiLan Proof-of-Concept randomized controlled trials [3,4,5].

LixiLan-L was a Phase III clinical trial that comprised a 6-week run-in with iGlar and a 30-week randomized treatment period comparing treatment with iGlarLixi (N = 366) versus iGlar (N = 365) in patients with T2D previously not sufficiently controlled on basal insulin with or without OADs. The primary analysis showed superior glycemic control as assessed by the change in HbA1c from baseline to Week 30. Furthermore, at Week 30, 54.9% (n = 201) of patients treated with iGlarLixi achieved the HbA1c target of < 7.0% compared with 29.6% (n = 108) of patients with iGlar alone (p < 0.0001) [3]. This post hoc analysis was designed to evaluate the impact of HbA1c levels measured at screening on glycemic control in 30-week completers of the LixiLan-L trial. The inclusion of the screening phase allowed the 6-week run-in with iGlar to be evaluated in conjunction with the 30-week treatment period, providing a more complete investigation of treatment during the LixiLan-L trial.

Methods

The present study is a post hoc analysis of the LixiLan-L trial (NCT02058160), the methods of which are described briefly below; the complete methodology has been described previously [3].

Trial Design

The LixiLan-L trial was a randomized, open-label, parallel-group, multinational, multicenter Phase III clinical trial for patients previously uncontrolled on basal insulin with or without OADs. The trial was initiated on January 27, 2014 and ended on July 9, 2015 and comprised an 8-week screening phase, which included up to 2 weeks of screening and a 6-week run-in, followed by a 30-week treatment period. The primary efficacy endpoint of the LixiLan-L trial was change in HbA1c from baseline to Week 30.

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964, as revised in 2013. Informed consent was obtained from all patients for being included in the study, and the publisher’s policy concerning informed consent was followed. The protocol also complied with the laws, regulations, and any applicable guidelines of the countries in which the study was conducted. Institutional review boards or independent ethics committees at each study site approved the LixiLan-L study protocol.

Data that could identify treatment were masked prior to data review and event adjudication during the LixiLan-L trial. A data monitoring committee reviewed and analyzed the safety data provided by an independent statistical group throughout the LixiLan-L trial.

Study Population

The LixiLan-L trial recruited outpatients with T2D. Eligibility requirements to participate included a diagnosis of T2D for at least 1 year, treatment with a stable basal insulin dose (15–40 U/day) and a FPG level of ≤ 180 mg/dL (10.0 mmol/L) at screening for patients on basal insulin and two OADs, or one OAD other than metformin, or an FPG of ≤ 200 mg/dL (11.1 mmol/L) for patients on basal insulin with or without metformin. The HbA1c level was required to be between 7.5 and 10%, inclusive.

Interventions

At run-in, all patients were on iGlar (100 U/mL), administered once daily; patients previously on other basal insulins were switched to iGlar at the beginning of the run-in. Treatment with OADs other than metformin was stopped. During the run-in, iGlar doses were adjusted according to investigator discretion to achieve a daily fasting self-monitored plasma glucose (SMPG) level of ≤ 140 mg/dL (7.8 mmol/L) while avoiding hypoglycemia.

At the end of the run-in, patients who had an HbA1c of  ≥ 7% and ≤ 10%, a mean fasting SMPG of ≤ 140 mg/dL (7.8 mmol/L), and an iGlar dose of ≥ 20 and ≤ 50 U/day were randomized (1:1) to receive iGlarLixi or iGlar, stratified by HbA1c at Week − 1 (< 8%, ≥ 8%) and metformin use at screening (Yes, No). Randomization was performed by an interactive voice response system/interactive web response system according to the randomization scheme provided by the study statistician.

iGlarLixi was provided in two prefilled SoloSTAR® pens (Sanofi, Paris, France). Pen A [2 U iGlar (100 U/mL)/1 µg Lixi ratio] was used to deliver iGlarLixi doses between 10 U (10 U/5 µg) and 40 U (40 U/20 µg); Pen B (3:1 ratio of iGlar:Lixi) was used to deliver iGlarLixi doses between 30 U (30 U/10 µg) and 60 U (60 U/20 µg). iGlar was provided in a prefilled Lantus® SoloSTAR® pen (100 U/mL; Sanofi, Bridgewater, NJ).

In order not to exceed the recommended Lixi starting dose of 10 µg/day, patients randomized to iGlarLixi started on Pen A at a dose of 20 U/day (20 U iGlar/10 µg Lixi) if previously on an iGlar dose of < 30 U/day or on Pen B at a dose of 30 U/day (30 U iGlar/10 µg Lixi) if previously on an iGlar dose of ≥ 30 U/day. iGlarLixi was self-administered prior to breakfast (within 60 min). The iGlarLixi dose was kept stable for 2 weeks. Patients receiving iGlar were started on the same daily dose of iGlar received the day before randomization. iGlar administration time was determined at the beginning of the run-in according to the patient’s and/or investigators’ preference and was at approximately the same time each day. During treatment, iGlarLixi and iGlar were titrated based on iGlar dose to a fasting SMPG of 80–100 mg/dL (4.4–5.6 mmol/L) while avoiding hypoglycemia.

Lifestyle and diet counseling was provided at the start of the run-in phase and at randomization and was to be continued during the study. Compliance with the diet and lifestyle counseling was to be assessed if sufficient glucose control was not achieved.

The need for rescue therapy was determined according to central laboratory-measured FPG and HbA1c (after Week 12) levels, which were measured if the patient’s recorded fasting SMPG values on three consecutive days exceeded the threshold limit for the corresponding period of the study (Electronic Supplementary Table S1).

Post Hoc Analysis

This post hoc analysis of the LixiLan-L trial assessed the efficacy of iGlarLixi compared with iGlar alone in 30-week completers from the modified intent-to-treat population according to HbA1c level at screening, including both the screening/run-in phase and the 30-week treatment period. Thirty-week completers were defined as patients who completed the 30-week treatment period without rescue therapy. Patients were split into three subcategories according to HbA1c level at screening: HbA1c ≤ 8%, 8% < HbA1c ≤ 9%, and HbA1c > 9%. The clinical endpoints measuring glycemic control included change in HbA1c from screening to Week 30 and the proportion of patients achieving an HbA1c target of < 7% at Week 30. The dose of iGlar and Lixi at Week 30 and documented symptomatic hypoglycemia during the treatment period, defined as an event with typical symptoms of hypoglycemia that were accompanied by a measured plasma glucose concentration of ≤ 70 mg/dL (≤ 3.9 mmol/L), were also evaluated.

Statistical Analyses

For the overall category, the least squares (LS) mean was estimated from an analysis of covariance (ANCOVA) model with treatment groups, randomization strata of HbA1c (< 8, ≥ 8%) at Week − 1, randomization strata of metformin use at screening (Yes, No), and country as fixed effects, and screening HbA1c value as a covariate. For the screening HbA1c subcategories, the LS mean was estimated from an analysis of variance (ANOVA) model with treatment groups, randomization strata of metformin use at screening (Yes, No), subgroup factor, treatment by subgroup factor, and country as fixed effects. The number (%) of patients with any documented symptomatic hypoglycemia during the 30-week treatment period, as well as the number of events per patient-year, were summarized by treatment and screening HbA1c subcategories.

Results

Patient Characteristics and Demographics

The overall group of 30-week completers comprised 660 patients who completed treatment with iGlarLixi (n = 327) or iGlar (n = 333) without rescue therapy (Table 1). Patient demographics and characteristics at screening and baseline for the 30-week completers were similar between treatment groups overall and within each HbA1c screening subcategory (Table 1).

Table 1 Demographics and characteristics at screening or baseline of the 30-week completers of the LixiLan-L trial (30-week completers from the modified intent-to-treat population)
Full size table

HbA1c Reduction

For the 30-week completers of the study, greater reductions in HbA1c from screening to study end was achieved with iGlarLixi than with iGlar (p < 0.0001) (Table 2; Fig. 1). The LS mean HbA1c change [± standard error (SE)] was − 1.7% ± 0.1 for iGlarLixi and − 1.1% ± 0.1 for iGlar, with an LS mean difference of − 0.5% ± 0.1 (p < 0.0001) for iGlarLixi versus iGlar. Only iGlarLixi-treated patients achieved a mean HbA1c of < 7% at Week 30.

Table 2 Subpopulation analyses at Week 30 for 30-week completers of the LixiLan-L trial (30-week completers from the modified intent-to-treat population)
Full size table
Fig. 1
figure 1

Mean glycated hemoglobin (HbA 1c ) change in 30-week completers of the LixiLan-L trial based on screening HbA1c values (30-week completers from the modified intent-to-treat population). All data are observed values. iGlar Insulin glargine U100, iGlarLixi Fixed-ratio combination of insulin glargine + lixisenatide

Full size image

Regardless of the HbA1c screening subcategory, reductions in HbA1c from screening to Week 30 were greater in patients receiving iGlarLixi than in those receiving iGlar (p < 0.0001 for all) (Table 2; Fig. 1) and allowed a higher proportion of patients (52.2–74.2 vs. 23.5–37.2%, respectively; Fig. 2) to reach the target HbA1c of < 7% at Week 30. The numerically largest change in HbA1c was observed with iGlarLixi in the HbA1c > 9% screening subcategory (LS mean change − 2.4%). In all HbA1c screening categories, a mean HbA1c of < 7% was only achieved with iGlarLixi.

Fig. 2
figure 2

Patients reaching the HbA1c < 7% target at Week 30 (30-week completers from the modified intent-to-treat population). aProportion difference = difference in the proportions of patients; weighted average of proportion difference between treatment groups from each strata [randomization strata of metformin use at screening (Yes, No)] using Cochran–Mantel–Haenszel weights. CI Confidence interval

Full size image

Treatment Dose

The final insulin dose at Week 30 was generally comparable between treatments, overall and for each HbA1c screening subcategory (Table 2). In the iGlarLixi treatment group, the corresponding final Lixi dose was approximately  17 μg, irrespective of HbA1c level at screening.

Hypoglycemia

In all 30-week completers, the incidence of documented symptomatic hypoglycemia was similar between those receiving iGlarLixi and those receiving iGlar during the 30-week treatment period (Fig. 3). In the subgroup with HbA1c of ≤ 8% at screening, the incidence of documented symptomatic hypoglycemia was numerically lower with iGlarLixi versus iGlar (36.1 vs. 47.9%). For the higher HbA1c screening subcategories the incidences were similar for iGlarLixi and iGlar. Documented symptomatic hypoglycemia events per patient-year were numerically lower with iGlarLixi versus iGlar overall and in all HbA1c screening subgroups, most prominently in the lowest HbA1c screening category (overall: 2.7 vs. 4.2; HbA1c ≤ 8%: 1.8 vs. 5.1; 8% < HbA1c ≤ 9%: 2.8 vs. 3.8; HbA1c > 9%: 3.7 vs. 4.2; Fig. 3).

Fig. 3
figure 3

Documented symptomatic hypoglycemia during the 30-week treatment period (30-week completers; modified intent-to-treat population). aDocumented symptomatic hypoglycemia includes events with typical symptoms of hypoglycemia and a measured plasma glucose concentration of ≤ 70 mg/dL (≤ 3.9 mmol/L). p-y Patient-year

Full size image

Discussion

In this post hoc analysis of 30-week completers from the LixiLan-L trial, iGlarLixi treatment led to patients achieving a mean HbA1c level of < 7% across all HbA1c screening subcategories and overall, meeting the ADA-recommended target. iGlarLixi was more effective than iGlar in controlling HbA1c across all subgroups, including those with a screening HbA1c level of > 9%, without increasing the risk of hypoglycemia. Subgroups with higher initial HbA1c values had the greatest reduction in HbA1c for both treatment strategies.

We investigated the HbA1c reductions achieved from screening to study end for 30-week completers in order to obtain a more complete picture of the treatment provided to the patients during the clinical trial. We also looked at the glycemic control achieved according to the specified screening HbA1c subcategories.

As reported previously for the LixiLan-L trial, mean HbA1c levels decreased from 8.5% at screening to 8.1% at randomization, followed by a LS mean reduction of − 1.1% to an HbA1c of 6.9% at study end for the iGlarLixi treatment group (modified intent-to-treat/mixed-effect model with repeated measures) [3]. We observed a similar mean HbA1c decrease for 30-week completers according to the present post hoc analysis (HbA1c: 8.5% at screening, 8.0% at baseline, and 6.9% at study end). In a prespecified analysis of the LixiLan-L trial, which analyzed the impact of baseline characteristics, HbA1c reductions from baseline to Week 30 were greater with iGlarLixi than with iGlar for both the baseline HbA1c < 8% and HbA1c ≥ 8% subcategories (p < 0.0001) [8]. Additionally, the higher baseline HbA1c subcategory (HbA1c ≥ 8%) demonstrated a greater HbA1c reduction compared with the lower HbA1c subcategory (< 8%) [8], similar to the trend shown here for the screening HbA1c subgroups.

Changes in body weight and lifestyle modifications during the study could also have impacted HbA1c reductions. In the primary analysis of data from the LixiLan-L trial, treatment with iGlarLixi resulted in a mean reduction in body weight from baseline (− 0.7 kg), whereas an increase in body weight (0.7 kg) was observed with iGlar at Week 30 (p < 0.0001) [3]. In the above-mentioned prespecified analysis of the LixiLan-L trial, the mean weight change was numerically different for patients between baseline body mass index (BMI) subgroups for both treatment arms (iGlarLixi: BMI < 30 kg/m2: − 0.1 kg vs. BMI ≥ 30 kg/m2: − 0.9 kg; iGlar: BMI < 30 kg/m2: 1.1 kg vs. BMI ≥ 30 kg/m2: 0.7 kg); however, the mean ± standard deviation change in HbA1c at Week 30 from baseline was comparable between BMI subgroups for both treatment arms (iGlarLixi: BMI < 30 kg/m2: −1.1 ± 0.9% vs. BMI ≥ 30 kg/m2: −1.1 ± 0.9%; iGlar: BMI < 30 kg/m2: −0.5 ± 0.9% vs. BMI ≥ 30 kg/m2: − 0.6 ± 0.9%) [8]. In addition, any effect on body weight as a result of the lifestyle and diet counseling provided before screening and during the study would have been applicable to both treatment arms.

Limitations of the LixiLan-L trial included the open-label study design and the relatively short 30-week study duration; longer trials are needed to assess the durability of the glycemic reductions observed [3]. Additionally, the post hoc approach of the present analysis may be considered to be a limitation.

Conclusion

In conclusion, irrespective of initial HbA1c screening levels, iGlarLixi can be considered to be an effective new treatment option for controlling HbA1c without an increased risk of hypoglycemia.